As some localization community members posted to dev-mdc newsgroup, the current Deki system has many and many system or feature regressions compared the past MediaWiki based system. Of course, we know that the Deki system was advertized as having some great features, such as a GUI interactive editor, or its highly customizable template system. But it seems clear that these features will NEVER been able to cover the regressions or the system difficulties. Only little localization contributor needed or did not want some of them especially for a GUI interactive editor.
So, we want to propose the following things to the heads or MDC-leads of the Mozilla Corporation.
1. Have more and more discussions with localization communities before any high-impact decision
Before the deki system change as well as for the switch over to the MDN, we only have heard the final decision from MoCo. For the MDN, the project page had ‘open’ on wiki.mo, but no official announces or discussions to/with the localization communities on such as dev-mdc newsgroup before such FINAL announcement.
For example, while we all agree that there were a few problems with MediaWiki related with the custom features, title-override and breadcrumbs. It should be widely known that many localization communities had argued against the switch to Deki, feeling in advance that it would only lead to more problems. (Remark: breadcrumbs’ feature had implemented to the MediaWiki after that.)
Also for MDN, to engage with the hacks, the AMO, the developer hub, we must make clear the issues related on localizations before every work. It is hard to catch-up or modify the built systems to be aware of the localization after the lunch, also each localization community would have their own sites on similar concepts to these which is better to be integrated to MDN.
2. Be aware of the localization communities
Many and many system requirements, most of whose are regressions, are proposed from the communities, but it seems that they are really not taken as seriously as they should be.
For example, the diff view is still not satisfactory more than two years after the switch.
3. Build a better wiki system
SUMO Project are now discussing to build their system from scratch — Kitsune, and we might be able to do some discussion with them for the MDC itself. Or we have another possibility — MediaWiki.
For this purpose, we have made a short list of issues related on localization work, and propose to make discussion among the localization communities for the desired requirements’ list to build and enlarge localization communities.
Co-Proponents or Co-sponsors of this proposal:
- himorin – on behalf of MDC Japan Project
- potappo – on behalf of MDC Japanese localization community
- Benoit – on behalf of MDC French localization community